A Coastal Storm Modeling System for determination of flood hazards along a high energy coast in response to SLR and 21 st century storms

Size: px

Start display at page:

Download "A Coastal Storm Modeling System for determination of flood hazards along a high energy coast in response to SLR and 21 st century storms"

David Johnson
5 years ago
Views:

1 Li Erikson, Patrick Barnard, Andrea O Neill, Jodi Eshleman, Amy Foxgrover, Michael Fitzgibbon 2, Grant Ballard 2, Kelly Higgason 3 A Coastal Storm Modeling System for determination of flood hazards along a high energy coast in response to SLR and 21 st century storms 3 2

Motivation & Background GOAL of the work: To assess the vulnerability of the coastal margin to flooding due to 21 st century sea level rise and coastal storms Objectives of the work: Emphasis on

2 Motivation & Background GOAL of the work: To assess the vulnerability of the coastal margin to flooding due to 21 st century sea level rise and coastal storms Objectives of the work: Emphasis on directly supporting federal and state climate change guidance and vulnerability assessments Location-independent methodology (widely applicable) Address all relevant contributions to total water levels and flooding SLR tides steric effects storm surge, waves river discharge levees and seawalls non-linear interactions

3 Presentation objectives elucidate on the added benefit (or not) of accounting for the influence of SLR on storm induced flood hazards along high-energy coastlines and provide an overview of web tools and the underlying CoSMoS model developed for evaluating future flood vulnerabilities

4 Talk outline 1. Overview of the end-user web tool and modeling approach 2. Example application to North-Central California 3. Findings flood levels are non-linearly related to increases in SLR accounting for storms in combination with SLR, substantially increases flood extents, but is strongly a reflection of the topography (duh, no surprise!)

5 1. Overview of the end-user tool and modeling approach

6 OCOF web tool

7 OCOF web tool

8 Generate summary reports of your area of interest

9 HERA web tool 2 m SLR year projected storm 355,000 residents $102 billion in property 1,941 miles of roads 278 critical facilities

10 web tools & underlying model Hazard Exposure Reporting and Analytics (HERA) Web tool for data visualization, synthesis, and download Web tool for socio-economic web

CoSMoS modeling approach Global Scale Deep

astronomic tides (Delft3D+SWAN) Local Scale

storm surge, tides, overland flow, fluvial

11 CoSMoS modeling approach Global Scale Deep water wave generation & propagation (WW3 & GCMs) Regional Scale Swell propagation, wave generation, storm surge, and astronomic tides (Delft3D+SWAN) Local Scale Nearshore waves, wave setup and runup, storm surge, tides, overland flow, fluvial discharge, long-term topobathy change (Delft3D+SWAN + XBEACH) 11

12 CoSMoS global-scale modeling WW3 (ver. 3.14, TC physics pack.) winds from 4 CMIP5 GCMs for simulation of deep water waves; historical, RCP4.5 & RCP 8.5 Bias compared to ERA-I (all months) (a) (b) 1 x 1.25 (c) (d) ENP 0.25 (e) Example wave model simulation

13 Regional & local-scale modeling 2DH Delft3D & SWAN 2D [1D] XBeach waves currents depths Waves, currents, WLs, event-based morphodynamic change

14 Why XBeach in model train? Rapid compuation of event-driven morphodynamic change Inclusion of infragravity (IG) wave energy Incident band is important to generate offshore transport and stir sediment Infragravity band required to help short waves reach the upper beach and dune, modulate strong offshore currents, and is often main contributor for overwash Both types required for accurate modelling

15 Why XBeach in model train? Two main types of IG waves bound infragravity waves generated offshore by, and travelling with, wave groups (generally dominant on shallow, dissipative beaches) breakpoint generated IG waves; created at the breakpoint of short waves (moving breakpoint mechanism; more important on steep beaches)

16 Why XBeach in model train? Short wave envelope IG, wave setup, SS, tides Courtesy: Robert McCall, Deltares

17 PNW ( ) 2. Example application to North-Central California

18 Delft3D and SWAN model grids # grid cells Res. (m 2 ) Tier 1 FLOW 157,112 2k to 4k Tier 2 FLOW north 560,368 9 to 688 Tier 2 FLOW south 342,019 6 to 980 Tier 2 WAVES north 96, to 611 Tier 2 WAVES south 98, to 725

19 XBeach model grids 933 profiles Topo and bathy extracted from 2m DEM 30m resolution offshore 5m resolution shoreward of the 2.5m water depth +10m Sub-areal profile sections with slopes > 32º considered to be immobile revetments or cliffs -10m 100m 2m seamless DEM gov/topobathy_viewer

20 Evaluation of model skill Pt. Reyes tide gauge CDIP 142 NDBC46026 SF tide gauge Cliff House camera FOV Cliff House Timestack imagery for runup obs.

21 3. Findings storm-related flood potentials are non-linearly related to SLR

the use of wave setup rather than wave runup is often preferred since the swash lens is often thin

22 Coastal flood elevation potentials CoSMoS TWL & ZFL depicted here shows linear super-position (terms are mutually exclusive) Total water levels, TWL = f SLR, η AT, η NTR, R For flood hazard analysis, the use of wave setup rather than wave runup is often preferred since the swash lens is often thin and contains a limited volume of seawater (e.g., Barnard et al. 2014), ZFL = f SLR, η AT, η NTR, η ҧ. 22

23 TWLs and ZFLs without SLR H o = 2.4m; T p = 13s 3.88m H o = 6.9m; T p = 17s H o = 9.8m; T p = 19s 7.45m 23

24 H s (m) TWLs and ZFLs without SLR H o = 2.4m; T p = 13s 3.88m H o = 6.9m; T p = 17s high H o = 9.8m; T p = 19s 7.45m low 24

25 TWLs and ZFLs without SLR 75 th percentile median 25 th percentile Box plot legend 100yr storm, 0 m SLR medians highest at beach-fronted infrastructure, but maximum values are fairly consistent for all back-beach types except sand spits where overwash occurs during storm events.

26 TWLs and ZFLs without SLR Wave setup ZFL η NTR + η AT + MSL 26

27 Non-linear flood potential w.r.t. SLR δ i i i = ZFL SLR SLR ZFL 000 Cliffs, bluffs and structures with no fronting beach are less prone to this error since waves approaching these types of configurations are likely to break close to or upon impact with the bluff or structure leaving little accommodation space for wave setup to develop. d is small for low SLRs but quite high for SLRs 1.5m particularly for beaches, dunes, and beachfronted cliffs and structures for which the 75 th percentile difference is ~60cm (linear superposition would under-estimate ZFL) 27

28 Non-linear flood potential w.r.t. SLR at least 2 likely reasons Greater water depths allow waves to reach close to shore before shoaling, refraction, and consequent energy dissipation. Assumed initial static profile combined with immediate SLR results in wave breaking and runup along a different sections of the profile. 28

29 3. Findings effect on flood extents

Importance of accounting for storms Sea level rise (m) Flooding increases dramatically with SLR Including storms increases flood extents by another 4%

30 Importance of accounting for storms Sea level rise (m) Flooding increases dramatically with SLR Including storms increases flood extents by another 4% (Muir) to 20% (Ocean Beach) The added contribution from storms is negligible at 5m SLR for 2 sites, a reflection of the steeper topography further inland

31 Flooded / inundated area (km 2 ) Importance in accounting for storms 6 5 SLR only SLR & storm cm 50 cm 100 cm 150 cm Sea level rise

32 Summary and Conclusions

33 Summary and conclusions CoSMoS and associated web tools developed with the aim of supporting federal and state climate change guidance and vulnerability assessments Aims to address all relevant contributions to total water levels and future flood hazards, considering SLR tides steric effects storm surge, waves river discharge levees and seawalls non-linear interactions Each of the components that contribute to ZFL and TWL are computed numerically including SLR effects on wave propagation and wave-current interactions, but at a high computation cost is it worth the computation cost?

34 Summary and conclusions Flood levels are shown to be non-linearly related to SLR, suggesting that simple linear superposition of static flood levels with SLR will results in under or over-estimates of flood hazards Non-linearity increases with SLR, and is most prominent for SLR>1.5m Non-linearity particularly evident at beaches, dunes, and beach fronted cliffs/structures (75 th percentiles deviate by ~0.6m from simple linear super-position) Cliffs, bluffs and structures with no fronting beach less prone to the non-linear response in Z to SLR

35 Summary and conclusions The non-linear response of Z to SLR, is in part due to 1. swell reaching the shore are greater compared to the no SLR case deeper nearshore waters (increased SLR) allow waves to reach closer to shore before loosing energy due to shoaling & refraction, (~0.05 SLR) changes in wave current interactions (increase swell by another 5% for a total of ~0.10 SLR) 2. and because raising SLR onto an assumed initial static profile enables wave breaking, setup, and runup to act along different section of the profile largely due to waves breaking close to or upon impact with bluff or structure leaving little accommodation space for wave setup to develop

36 Limitations, uncertainties & future steps assumes an unchanged initial profile future profiles will have evolved with SLR and in areas with infrastructure, coastal squeeze may significantly alter the profile, increasing the potential flood vulnerability Local wind-wave growth is not accounted for in these results potential to increase the d discrepancy (under-estimate of linear super-position) seas have been computed in all other CoSMoS simulations, but have yet to be analyzed for d

37 Thank you. Questions?

City of Del Mar Local Coastal Plan (LCP) Amendment for Sea Level Rise and Coastal Flooding

City of Del Mar Local Coastal Plan (LCP) Amendment for Sea Level Rise and Coastal Flooding 6/22/17 Presentation to Sea Level Rise Stakeholder- Technical Advisory Committee (STAC) Nick Garrity, P.E. Lindsey